Contribute for an Ontology for Smart Homes and their Conversational Assistants

An Accessible Smart Home Based on Integrated Multimodal Interaction

“…Next, a study with twenty participants was carried out [ 12 ], providing more important capabilities to consider: remote control, home state report, spoken communication, querying and control of appliances and lights, activity scheduling for appliances, control of the temperature for appliances (e.g., water heater) and home divisions, and obtaining consumption information. A new version of the assistant was implemented considering these results, which was then evaluated by six participants, who carried out several tasks using the assistant together with a simulated home [ 28 ].…”

“…To enable an integrated view of the home, we adopted Design for Data and a semantic-based approach for all the information [ 1 , 12 ]. This approach allows the storage of all relevant information on the home and its devices in an organized way, as well as enables richer queries from the users.…”

“…More specifically, all data provided by the home devices are stored in the home database, relying on the domain ontology proposed by our group [ 1 , 12 ], and information is retrieved from the database using SPARQL queries. The ontology includes the following classes: Partition, for describing home divisions; Device, for home appliances and devices present in each home division; Resource, representing resources consumed by each device, such as water, electricity, or gas; Consumption, representing the consumption of a given resource by a device; Error, for errors associated with a device; and Issue, to describe the occurrence of a specific error for a given device.…”

“…Assistants have the potential of providing a natural and integrated way of interacting with smart homes. However, despite their popularity and recent advancements, commercially available home assistants still have some limitations, for example, they tend not to provide a unified view on the home, i.e., they do not store data regarding the home devices in a structured way, not allowing to perform rich complex queries for obtaining relevant information (e.g., consumption of a resource by all devices in a given home division) [ 12 ]. Multi-turn conversation is also either lacking or needs to be improved to become more natural and engaging [ 13 , 14 ].…”

“…Next generations of home assistants need to contemplate new capabilities and be more aligned with users’ needs. A recent inquiry [ 12 ], by the authors, to 20 participants aged between 10 and 63 years old, with different backgrounds, revealed some capabilities deemed important by users: home state report, temperature control of appliances (e.g., water heater, oven), and information regarding resources consumption.…”